| dc.description.abstract | (i) The history of biochemistry is studded with many landmarks, and one of the significant events is the discovery and characterization of vitamins. Folic acid, which was isolated from spinach leaves for the first time, was shown to be an anti anemia factor. This vitamin, in its coenzyme forms (Fig. 1), functions as a carrier of one carbon (C1) fragments required for the biosynthesis of methionine, thymidylate, purines, tRNA, and a large number of end products that arise by the transfer of a methyl group from S adenosylmethionine (Fig. 2).
(ii) The three enzymes involved in the “thymidylate cycle”-namely SHMT (serine hydroxymethyltransferase), DHFR (dihydrofolate reductase), and TS (thymidylate synthase)-from different species show a considerable degree of evolutionary conservation in their active site peptides. A brief survey of these enzymes isolated and characterized from a wide variety of sources is presented.
(iii) SHMT resembles other PLP dependent (pyridoxal 5 phosphate) enzymes in its mechanism of action and active site structure. It also has several characteristic features that distinguish it from other PLP enzymes. SHMT can be considered the first enzyme in the pathway for the interconversion of folate coenzymes and is the source of C1 units in mammalian systems. Recent studies revealed that the levels of SHMT as well as its kinetic properties are altered in neoplastic tissues. These results suggest that this enzyme could be exploited as an alternate target in cancer chemotherapy.
(iv) The objectives of the present investigation were:
(a) to isolate SHMT from normal human liver in a homogeneous form, characterize the interactions of the human liver enzyme with effectors, and correlate ligand and denaturant induced conformational changes with catalytic activity;
(b) to probe the nature of the active site using dyes and specific reagents that modify amino acid residues present at the active site;
(c) to confirm the regulatory nature of the enzyme; and
(d) to compare the properties of the enzyme in normal and neoplastic tissues.
(v) The procedures employed include: SHMT activity assays; chemical, enzymatic, and spectral estimation of H4 folate (tetrahydrofolate; THF); protein purification techniques; electrophoretic and immunological methods; chemical modification of active site residues; amino acid analysis; and absorbance, circular dichroism (CD), and fluorescence spectroscopy.
(vi) SHMT from human liver was isolated by (NH ) SO fractionation, adsorption and elution from CM Sephadex, gel filtration on Sephacryl S 200, and affinity chromatography on Blue Sepharose (Table 7). The specific activity of the purified enzyme was in the range 3.6–3.9 units. The enzyme was homogeneous as determined by PAGE (Fig. 14A), SDS PAGE (Fig. 14B), gel filtration, immunodiffusion (Fig. 15A), and immunoelectrophoresis (Fig. 15B). The enzyme had a molecular weight of 220,000 and was a homotetramer with a subunit molecular weight of 55,000 (Figs. 16, 17).
(vii) Reaction rates were linear up to 1.6 µg enzyme per 0.1 mL reaction mixture and up to 15 min of assay (Fig. 18A,B). The enzyme functioned optimally at 50 °C (Fig. 18D), and an activation energy of 10 kcal·mol ¹ was calculated from the Arrhenius plot (Fig. 18B). A pH optimum of 7.5 (Fig. 18C) and a catalytic center activity (at 37 °C and pH 7.4) of 200 min ¹ were obtained. The PLP bound to the holoenzyme was dissociated by treatment with L cysteine and ammonium sulfate, followed by dialysis (Fig. 19; Table 8). The apoenzyme was inactive and was fully reconstituted with PLP (Table 8). A reducing agent such as 2 ME (2 mercaptoethanol) was required to protect the enzyme against inactivation on storage.
(viii) The enzyme had visible absorbance and CD peaks at 425 nm and 422 nm, respectively, due to bound PLP (Figs. 19, 20). The far UV CD spectrum showed negative ellipticity bands at 210 and 220 nm (Fig. 21). The helical content was calculated to be ~20%. L Ser and H4 folate caused significant quenching of the visible CD spectrum (Fig. 30), indicating a changed environment of the bound PLP at the active site upon interaction with L Ser and H4 folate.
(ix) Aromatic residues in the enzyme gave a characteristic fluorescence emission maximum at 335 nm when excited at 280 nm (Fig. 22).
(x) The human liver enzyme purified by this procedure resembled the enzyme from other sources in its physicochemical and catalytic properties and exhibited the general spectral and chemical properties of PLP enzymes.
(xi) Saturation of the enzyme with L serine was hyperbolic, with an apparent K of 1.6 mM (Fig. 24).
(xii) Evidence for multiple binding sites on the enzyme was provided by partial inhibition produced by H4 folate, MTX (methotrexate), O phosphoserine, while 5 CHO H4 folate (5 formyl THF), dichloro MTX, O acetylserine, and O methylserine were complete inhibitors (Figs. 25, 26, 29).
(xiii) The H4 folate saturation of the enzyme was sigmoidal, with nH = 2.6 and an apparent K of 0.9 mM (Figs. 31, 32). Preincubation with L Ser reduced the cooperativity of folate interaction to nH = 1.1 (Table 9). NAD and NADH were negative and positive allosteric effectors, respectively (Table 9; Figs. 31, 32).
(xiv) Recently, the homotropic cooperative interactions of H4 folate with SHMT were attributed to oxidation of H4 folate in the assay at low concentrations. A major difference between the studies demonstrating cooperativity and those criticizing these results was the range of H4 folate concentrations used to determine the saturation pattern (i.e., 0.1–2.7 mM in the former vs 0.02–0.2 mM in the latter). In addition, studies demonstrating cooperativity used a radioactive assay, whereas the other study used a coupled spectrophotometric assay.
(xv) Oxidation of H4 folate at one of the lowest concentrations used in this thesis (0.25 mM) was ruled out by spectral, chemical, and enzymatic estimation of H4 folate after incubation under assay conditions (Table 10; Fig. 33). The H4 folate saturation pattern was independent of incubation time (1–15 min) for measuring velocity (Table 11) and was unaltered under nitrogen (Fig. 34). These results rule out oxidation of H4 folate as the cause of the observed cooperativity.
(xvi) The nH value of SHMT varies from 1.0–3.9 (p. 106), depending on the enzyme source, and the observation that sigmoidicity can be altered by specific effectors (Table 9) emphasizes the specificity of these interactions. It is concluded that cooperative interactions of the enzyme with H4 folate are an inherent property of the enzyme and not an assay artifact.
(xvii) The enzyme was heat stable, with the midpoint of thermal stability curves at 55 °C (Fig. 40). The Arrhenius plot of log V vs 1/T (Fig. 18E) showed a break at 25 °C, followed by a linear phase up to 50 °C, and a denaturation phase above 50 °C. Activation energies of 10 kcal·mol ¹ (0–25 °C) and 4 kcal·mol ¹ (25–50 °C) were calculated from the slopes (Fig. 18E).
(xviii) L Serine protected the enzyme against heat inactivation and the loss of secondary structure caused by thermal denaturation. Denaturation was further examined by monitoring structural changes at extreme pH via [ ] (far UV CD), enzyme activity, and fluorescence at 335 nm. All three had maximal values at pH 7.4–7.5 (Figs. 18C, 43, 44, 45).
(xix) Addition of urea or GdmCl (guanidinium chloride) decreased the mean residue ellipticity at 222 nm in the far UV CD spectrum (Figs. 49, 50). The decreases in [ ] and enzyme activity followed identical patterns when expressed as a function of urea concentration (Fig. 52B); such parallelism was absent with GdmCl-activity was lost earlier than secondary structure (Fig. 52A). Urea and GdmCl also perturbed the fluorescence spectrum (Fig. 51), indicating conformational flexibility in human liver SHMT.
(xx) Cibacron Blue (CB) completely inhibited enzyme activity (Fig. 53), and NADH and H4 folate protected significantly against CB inhibition (Table 12). The inhibition was non competitive (Ki = 5 µM) with respect to L serine (Fig. 54).
(xxi) The dye interaction was further probed by monitoring characteristic difference spectra upon dye binding. The dye difference spectrum in the presence of enzyme showed an absorbance maximum at 675 nm and a trough at 590 nm (Fig. 55). The A at 675 nm increased with dye concentration (Figs. 56, 57). From A675 vs dye concentration, a binding constant of 5.0 µM was calculated using the Thompson & Stellwagen approach (Fig. 57). Using intrinsic protein fluorescence quenching (Fig. 62), a binding constant of 8 µM and a single binding site were evaluated from a Stern–Volmer plot (Fig. 63). Reversal of spectral changes upon addition of NADH and MTX (Table 13) suggests the dye binding site overlaps the folate and NADH binding sites.
(xxii) The red shift in the difference spectrum (Fig. 55) suggested dye binding in a hydrophobic pocket of the enzyme, confirmed by Figs. 58–60. These results indicate that hydrophobic interactions of the dye with the enzyme are strengthened by electrostatic forces (Table 13).
(xxiii) Chemical modification using phenylglyoxal (Arg; Fig. 70), DEPC (His; Fig. 76), and NBS (Trp; Fig. 79) indicated that at least one residue each of Arg, His, and Trp is essential for activity. Differential protection by substrates and effectors in modification experiments (Figs. 70, 71, 76, 77, 79; Table 18) suggests these residues occur at the active site.
(xxiv) The pH independence of binding of the substrate amino acid carboxylate suggests that Arg may bind this group at the active site. Arg may also bind the carboxyl groups of H4 folate’s glutamate tail, as indicated by protection by H4 folate (Fig. 71). The His residue probably functions as a general base in catalysis. The hydrophobic domain implicated by dye binding (Fig. 55) and the presence of Trp at active sites of some dehydrogenases support a Trp residue’s role in these properties.
(xxv) A plausible model (Figs. 93, 94) for substrate interaction with active site residues is proposed. H4 folate interacts within a hydrophobic pocket containing Trp and via electrostatic interactions between the glutamate carboxyls and positively charged Arg residues on the enzyme.
(xxvi) L Cysteine inhibited the enzyme completely (Fig. 82) and competitively with respect to serine (Fig. 83) with Ki = 33 nM (Fig. 83 inset). The apoenzyme prepared by L Cys treatment was fully reactivated by PLP reconstitution (Table 8; Figs. 81, 84). Addition of L Cys decreased the visible CD peak at 422 nm (Fig. 84). The first order rate constant for the L Cys interaction, calculated from CD change at 422 nm vs time (Fig. 85), was 1.0 × 10 ³ s ¹.
(xxvii) The kinetic mechanism for the interaction of DCS (D cycloserine) with SHMT was established by measuring changes in activity (Fig. 87), absorbance (Fig. 86), and CD spectra (Fig. 88). The interaction comprised an initial rapid step followed by two successive steps with rate constants of 6.0 × 10 ³ s ¹ and 1.08 × 10 s ¹ (Figs. 87, 88). By analogy with observations on sheep liver SHMT, a minimal kinetic mechanism for binding and inactivation of human liver SHMT was proposed (Fig. 96).
(xxviii) The enzyme exhibited altered kinetic properties in neoplastic tissues, with nH = 1.2 ± 0.2 (Tables 21, 23). A factor (nature not yet identified) appeared responsible for this kinetic change (Figs. 99, 100). The enzyme showed similar H4 folate saturation patterns in human and rat tumors (Figs. 99, 100; Table 23), and hyperbolic saturation in human fetal livers (Table 22).
(xxix) In conclusion:
(a) Human liver SHMT is a regulatory protein, exhibiting both positive and negative heterotropic interactions with substrates and effectors;
(b) the allosteric and active sites are likely proximal, with multiple interacting sites for folate and serine analogs;
(c) the enzyme is highly flexible, as shown by spectropolarimetric, fluorescence, and conformational studies correlated with catalytic function;
(d) Cibacron Blue binds in a hydrophobic pocket and likely overlaps H4 folate and nucleotide binding sites;
(e) Arg, His, and Trp are essential active site residues necessary for substrate binding and catalysis;
(f) DCS and L Cys inactivate the enzyme by converting it to the apoenzyme;
(g) SHMT in neoplastic tissues shows an altered regulatory pattern, suggesting a critical role in the supply of C1 units in neoplastic situations. | |
